Effect of stripping thyroxin from thyroxin-binding globulin on the measurement of free thyroxin in serum by equilibrium dialysis and by radioimmunoassay.

In considering factors that might influence measurement of free thyroxin (T4), we evaluated the proportion (%) of T4 that could be stripped from thyroxin-binding globulin (TBG). The percentage of free T4 was measured in serially diluted sera from four normal subjects, four patients with hyperthyroidism or hypothyroidism, four pregnant women, and four malnourished subjects with low TBG. The critical percentage of stripping was determined by the product of the percentage free T4 and the critical dilution factor (the point where the percentage free T4 began to decrease). The mean values obtained for the respective patient groups--6.38%, 2.76%, 16.73%, 9.75%, and 4.28%--were proved to be related to the rate of saturation of TBG with T4. Values for percentage stripping determined with the "GammaCoat two-step RIA" and the "LiquiSol RIA" were well within the critical percentage stripping by equilibrium dialysis, except in the case of low-TBG serum as measured by LiquiSol RIA. Free T4 concentration as measured by LiquiSol RIA decreased as sample volume decreased. These findings were ascribed to the relatively high values for percentage stripping in the LiquiSol RIA, which led to erroneously low values for free T4.